Improved Spatiotemporal Framework for Human Activity Recognition in Smart Environment.

The rapid development of microsystems technology with the availability of various machine learning algorithms facilitates human activity recognition (HAR) and localization by low-cost and low-complexity systems in various applications related to industry 4.0, healthcare, ambient assisted living as well as tracking and navigation tasks. Previous work, which provided a spatiotemporal framework for HAR by fusing sensor data generated from an inertial measurement unit (IMU) with data obtained by an RGB photodiode for visible light sensing (VLS), already demonstrated promising results for real-time HAR and room identification. Based on these results, we extended the system by applying feature extraction methods of the time and frequency domain to improve considerably the correct determination of common human activities in industrial scenarios in combination with room localization. This increases the correct detection of activities to over 90% accuracy. Furthermore, it is demonstrated that this solution is applicable to real-world operating conditions in ambient light.

Identification and Speed Estimation of a Moving Object in an Indoor Application Based on Visible Light Sensing of Retroreflective Foils.

Identification and sensing are two of the main tasks a wireless sensor node has to perform in an Internet of Things (IoT) environment. Placing active powered nodes on objects is the most usual approach for the fulfillment of these functions. With the expected massive increase of connected things, there are several issues on the horizon that hamper the further deployment of this approach in an energy efficient, sustainable way, like the usage of environmentally hazardous batteries or accumulators, as well as the required electrical energy for their operation. In this work, we propose a novel approach for performing the tasks of identification and sensing, applying visible light sensing (VLS) based on light emitting diode (LED) illumination and utilizing retroreflective foils mounted on a moving object. This low cost hardware is combined with a self-developed, low complex software algorithm with minimal training effort. Our results show that successful identification and sensing of the speed of a moving object can be achieved with a correct estimation rate of 99.92%. The used foils are commercially available and pose no threat to the environment and there is no need for active sensors on the moving object and no requirement of wireless radio frequency communication. All of this is achievable whilst undisturbed illumination is still provided.